| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Animal Health Centre, Morrinsville, New Zealand
EpiCentre, Institute of Veterinary, Animal and Biomedical Sciences, Massey University, New Zealand
1 Corresponding author: ccompton{at}ahc.co.nz
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: mastitis • epidemiology • heifer • peripartum
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Milk yield losses reported in heifers diagnosed with peripartum CM are variable, from less than 1% (Myllys and Rautala, 1995; Barnouin and Chassagne, 2001) to 5% (Oltenacu and Ekesbo, 1994) over a lactation, and 2.5 kg/d in the 7 d following Streptococcus spp. cases in the first week of lactation (Grohn et al., 2004).
First-calving heifers represent a valuable current and future resource. They make up the largest parity group in most herds, usually have the highest genetic merit of any age group in the herd, and, until a calf or milk is sold following their first calving, have not generated any revenues. For these reasons, diseases occurring at high frequency, and adversely affecting the production and lifetime performance of heifers must be a serious concern of dairy producers.
Little information has been published on the epidemiology of peripartum mastitis in pasture-grazed dairy heifers. Pankey et al. (1996) reported that 35% of heifers in pasture-grazed herds in New Zealand had one or more quarters diagnosed with IMI within 5 d following calving, and 8% of heifers had CM in the same period. But there are no data on the prevalence of IMI before calving in heifers in pasture-grazing farming systems, or on any productivity effects following naturally occurring IMI pre- or postcalving, or following CM in these systems.
Hence, the main aims of this study were to 1) describe at the quarter and heifer level the prevalence of IMI several weeks before, and within 5 d following calving, and 2) describe the incidence of CM in the first 14 d of lactation in first-calving heifers in pasture-grazed dairy herds. The study aimed to describe the bacteria involved in heifer peripartum IMI and CM and the repeatability of bacterial isolations over time. Additional aims were to estimate the risk of thelitis and loss of quarter symmetry or function by mid lactation, individual SCC (ISCC), milk yield and milk solids production at first postpartum production recording and averaged over the lactation, and the risk of premature culling in heifers.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Sample and Data Collection
At the time of enrollment, a single sample of mammary secretion was collected from each gland for bacteriologic examination following aseptic preparation of the teat end (only a single sample could be taken because of the low volume of secretion available). A commercial iodine-based teat antiseptic with 0.5% available iodine was applied by spray to the teats immediately after sampling. Duplicate milk samples were collected using the same method from all glands of each heifer within 5 d of calving during preplanned twice-weekly visits to the herds by trained technicians. If CM was diagnosed before a planned visit, duplicate milk samples were taken from all glands by a trained technician. Duplicate milk samples were collected from all first cases of CM occurring after the preplanned 1 to 5 d period; that is, between 6 and 14 d of lactation. On each sampling occasion, any abnormalities of the glands or teats were recorded. Heifer data including breed, New Zealand EBV, calving date, and individual animal milk production records were obtained electronically from a database (Livestock Improvement Corporation, Hamilton, New Zealand). All individual animal disease treatments from 1 mo before enrollment date and reason for removal of any cows from the herds throughout the lactation were collected from farm records. Results of microbiological tests of milk from heifers treated with systemic or intramammary antibiotics in the 21 d preceding sampling were excluded from analysis. At approximately 3 mo after the start of the calving period, each heifer was examined for the presence of thelitis (defined as a manually detectable thickening of the teat canal) and for the presence of nonfunctional or scantily functional mammary glands (defined as a visually apparent smaller gland compared with the contralateral gland in the same heifer immediately before attachment of milking units).
Bacteriological Examination
Microbiological procedures, diagnosis of IMI, and categorization of results were undertaken using standard methodology (Hogan et al., 1999). Milk samples were mixed thoroughly by inverting 2 to 3 times and then 10 µL was streaked onto a quarter plate of 5% sheep blood, 0.1% esculin agar (Fort Richard, Auckland, New Zealand) using a sterile disposable loop. Plates were incubated at 37°C for 48 h before reading results. All gram-positive, catalase-negative cocci were categorized as streptococci and further speciated by their esculin reaction, then CAMP (Christie, Atkins, Munch-Peterson) test results. Gram-positive, catalase-positive cocci were coagulase tested using a commercial kit (BBL Staphyloslide Latex test, Becton Dickinson, Sparks, MD) and categorized as either CNS or coagulase-positive (assumed to be Staphylococcus aureus). Gram-negative rods that could be identified with the basic biochemical tests (lactose, oxidase, triple sugar iron, Simmons citrate, and motility) were identified and recorded, and unidentified organisms were recorded as gram-negative rods. Gram-positive rods that could be identified with simple procedures were identified and recorded (e.g., Corynebacterium spp.). Bacillus organisms were identified by morphology only and recorded as Bacillus spp.; any unidentified organisms in this group were recorded as gram-positive rods. The number of colonies of each colony type was counted, up to a maximum of 50. Samples with more than 2 colony types were defined as contaminated. Samples from which fewer than 3 colonies of any 1 type of organism were found were recorded as a no growth, except for Staph. aureus where 
1 colony was recorded as an isolate. When duplicate samples were collected, both samples were required to have >2 colonies of the same bacterial species for the glands to be defined as infected. If 1 of the duplicate samples was contaminated, the results from the uncontaminated duplicate alone were used to diagnose infection.
Data Handling
Bacterial isolates were categorized as either major or minor pathogens. Bacterial species classified as major pathogens were Enterococcus spp., Escherichia coli, Klebsiella spp., Pasteurella spp., Proteus spp., Pseudomonas spp., Staph. aureus, Strep. agalactiae, Strep. dysgalactiae, and Strep. uberis. Minor pathogens were CNS, Corynebacterium spp., undifferentiated gram-negative rods, undifferentiated gram-positive rods, and yeasts. Bacteriological results from cases of CM that occurred within 0 to 5 d following calving were defined as IMI. Thus, the quarters defined as having an IMI also include those that were diagnosed with CM (this occurred in 195 quarters in 163 heifers). Clinical mastitis quarters were evaluated separately. Where a second pathogen was isolated from the gland, it was recorded as "isolate 2." Results of bacteriological testing of milk samples on one occasion and for repeated sampling were summarized at the quarter-level according to the definitions in Table 1
. When a quarter had both a major and a minor pathogen isolated at the same time, the quarter was given major pathogen status for that sampling occasion (n = 46 quarters precalving and 55 quarters postcalving), and the major pathogen was reported. When describing bacteriological results from samples taken at different occasions, the individual quarter bacterial isolates were compared. More than 1 bacterial isolate was identified from some quarters, and therefore, the categories for bacteriological status between samplings for a quarter were not mutually exclusive. For example, a quarter could be classified as both IMI "same" and IMI "new" if 1 of the precalving bacterial isolates was present at both pre- and postcalving samples (e.g., Staph. aureus) and a new bacteria was isolated postcalving (e.g., E. coli). Results of bacteriological testing of milk samples were aggregated from quarter to heifer level using the same major or minor pathogen quarter definitions. Results of heifer-level data were aggregated to the herd level to calculate prevalence of heifers within herds with IMI status before and within 5 d following calving, and within-herd cumulative incidence of CM within 14 d of calving. Results from samples and measurements taken >5 d postpartum that were for the routine postpartum sampling >14 d for CM cases were discarded for analysis. Breed of heifer was defined as the predominant parentage breed if greater than 11/16th (Friesian or Jersey), and all other breeds including crossbreds were classified as "other." In calculating proportions, samples that were either contaminated or not collected were not counted in the denominator. Records of gland function and presence or absence of thelitis at 3 mo after the start of the calving program were not analyzed from quarters that did not have a sample collected within 5 d following calving or were recorded with thelitis before or within 5 d of calving. Individual herd test records from heifers <10 d from date of calving or within 14 d of diagnosis of CM were excluded from analysis of ISCC.
|
2 test for trend in proportions (prevalence) of IMI relative to day of calving. The crude daily hazard (risk) of diagnosis of first case of CM within 14 d following calving was plotted separately for heifers and all other parity groups combined, and differences between survival curves for groups tested by log-rank test. Data from this study were of a hierarchical nature (quarters nested within heifers, in turn nested within herds), and observations at the lower 2 levels of measurement could not be considered independent of others within the same level. Analysis of such data with methods accounting for correlation between outcomes was necessary to avoid overly optimistic interpretations of probability values for association and biased point estimates (Dohoo et al., 2003). Hence, for dichotomous outcomes (reduced quarter size or function, thelitis, ISCC >200,000 cells/mL at herd tests, and risk of premature removal from the herd), linear mixed logistic regression models were fitted with multivariate normal random effects and random intercepts by penalized quasi-likelihood run within R (R Development Core Team, 2005). Continuous outcome measures (milk yield and milk solids production at the first herd test and average of 3 to 4 herd tests in lactation) were modeled using REML methods in the same software. For outcome measures recorded at the quarter level (e.g., quarter function or thelitis), 3-level models were used with both heifer and herd as random effects, and for those models with outcomes recorded at the heifer level (e.g., milk production, ISCC, and risk of removal), 2-level models were used with herd as a random effect. The choice of correlation structures for errors was made using biological reasoning and likelihood ratio test for improvement in model fit. Thus, autoregressive type 1 structure was used for repeated measures of continuous outcomes (e.g., milk production measures and log ISCC) and compound symmetry used for binary outcomes (e.g., quarters within heifers) or continuous measures made only once per heifer (milk production measures at first production recording in lactation).
Three-level generalized linear mixed regression models may be represented as
 |
where (g) refers to the link function (logit for logistic models), Yijk is the outcome variable, ßs are the model coefficients, Xs are the variables included in the models (prior bacteriological status and days from calving to sampling when postcalving IMI status used as a covariate), j refers to the heifer, k refers to the herd, and i to the ith quarter in the jth heifer in the kth herd, and the random effects are independent and normally distributed: µ heifer (j) 
N(0, 
2heifer), µherd (k) N(0, 2herd), 
ijk N(0, 2).
Heifer-level responses were modeled at 2 levels where the residual error was heifer and herd was a random effect. For production outcome models, predictor variables included were postcalving IMI and CM status (1/0), breed, days from calving to test date, and EBV as fixed effects. Individual SCC were natural log-transformed for analysis, and then back transformed for display of results. The variable "days from calving to test" was fit as single continuous covariate for models of measures of milk yield and milk solids production, because a polynomial term did not improve model fit. Nevertheless, a quadratic term was used for modeling average-lactation ISCC. Plots of residuals from all final models were examined to detect unusual patterns, but none were detected.
The intraclass correlation coefficient (ICC) estimates the resemblance among observations within a class (cluster) and may vary between 0 (independent) and 1 (totally correlated). It provides an indication of the infectiousness of an organism or similarity in susceptibility to disease because of common factors within the cluster and is required for estimating sample size in cluster surveys. Estimates of the ICC for a particular outcome at different levels of aggregation reflect the contribution of those levels to the overall variance, with the expectation that interventions targeted at a particular level with the highest ICC might have the greatest impact on the outcome (Dohoo et al., 2001). For binary data, a maximum likelihood method for point estimate and delta method for confidence intervals were calculated for each level of clustering using the method reported by Zou and Donner (2004).
Initially, unconditional associations between dichotomous outcomes and explanatory variables were examined using crude hazard ratios and incidence risk ratios for number of events, all evaluating categorical risk factors. Univariate ordinary logistic regression was used for continuous risk factors. Variables with associations significant at probability values 
0.20 were considered for inclusion in multivariable models. Wald tests were used to assess the significance of adding fixed effect terms in a forward stepwise way, and those with Wald test P-values 0.05 and variables known or suspected a priori as confounders (e.g., breed, DIM at milk test, and genetic breeding worth) were included in the final model. Other variables were considered for evidence of confounding and included in a final model if they caused >10% change in a regression coefficient due to their inclusion in the model, but there were none. All first-order interaction terms were tested in a forward stepwise manner and considered for the final model if significant (P 0.05), but none were found.
Because most outcomes were not rare, the odds ratio as measure of association could not be interpreted as multiplicative measures of risk and was not considered appropriate. Instead, incidence risk ratios (IRR) were used because they accurately describe the multiplicative risk of an outcome occurring for a given level of exposure compared with a reference exposure level. Incidence risk ratios and their confidence intervals were not obtainable directly from standard logistic regression models using the canonical logistic link, but were instead estimated from mixed logistic regression models using the log link (McNutt et al., 2003). Predicted population average estimates of continuous production outcomes were estimated from final models, using the most frequent categorical variable (Friesian breed) and the mean values of the other covariates as predictors.
Data were recorded in an Access (Microsoft Corp., Redmond, WA) database. Statistical significance for tests was declared at P 0.05, and confidence intervals reported for a 95% of range of values.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Quarter-Level Microbiological Results
Precalving quarter samples were taken 41 ± 16.4 d before individual calving date. The prevalence of infected glands precalving was 18.5% (Table 2), with CNS (13.5%) and Strep. uberis (2.8%) being the most prevalent isolates. Other bacterial isolates were infrequent and grouped together as "other." Quarter IMI prevalence precalving was significantly (P < 0.01) higher for minor compared with major pathogens. Dual infections in quarters precalving were uncommon; <2% of quarters had 2 different isolates identified in the same quarter. Forty of 47 dual infections were in combination with Strep. uberis, with a small number each of other possible combinations.
|
). For quarters with paired samples pre- and postcalving, absolute prevalence increased by 2.9%. This change was due to a significant increase in major pathogen prevalence (particularly Strep. uberis) from 3.9 to 11.6% and a 4.7% decline in minor pathogen prevalence (mainly CNS). The postcalving prevalence of major pathogens was significantly higher (P = 0.05) than that of minor pathogens. Other pathogens were infrequently isolated, and dual isolates were recovered from only 2% of quarters involving principally minor pathogens isolated in conjunction with Strep. uberis or Strep. dysgalactiae.
Minor pathogen prevalence was consistently higher than major pathogen prevalence precalving (P < 0.05 for all periods) and increased approaching calving (P < 0.01; Figure 1). Major pathogen prevalence remained low and did not differ significantly over the precalving sampling period (P = 0.38). Prevalence of minor pathogen IMI did not change for the specific interval between –27 and –9 d precalving and the day of calving (difference = 1.9%), but major pathogen prevalence increased [difference = 21.4%, confidence interval (CI) for difference: 16.6 to 26.5%] for the same period. Following calving, both major and minor pathogen prevalence declined rapidly after day of calving to <7% by d 5 following calving (P < 0.01 for both pathogen categories), and there was no difference in prevalence between pathogen categories within each period (P > 0.20).
|
Paired bacteriological results from the same quarters were compared over time after data was coded according to the definitions in Table 1. Overall, 1,755 of 2,530 quarters (69.9%) had samples pre- and postcalving that were both negative (Table 3). Of all quarters with pathogens isolated precalving, 35.8% had the same bacteria isolated postcalving; conversely, then, 64.2% of isolates self-eliminated. Yet, significantly more major pathogens (mainly Strep. uberis) than minor pathogens (mainly CNS) persisted over this period. Almost all (46 of 48 or 96%) quarters diagnosed with CM between 6 to 14 d following calving had the same isolate diagnosed at a prior sampling within 5 d of calving. Four hundred seventy-three quarters had IMI postcalving that were subclinical at the time of sampling, and of these 48 (10.1%) were subsequently diagnosed as CM within 14 d following calving.
|
). Postcalving, there was a higher prevalence of major (P < 0.01), but not minor pathogens, in rear vs. front quarters, and CM was more frequently (P < 0.01) diagnosed in rear compared with front quarters.
|
Clinical mastitis was diagnosed in 163 of 696 heifers (cumulative incidence = 0.234) within 14 d following calving. One hundred forty (20%), 17 (3%), 3 (<1%), and 3 (<1%) heifers had 1, 2, 3, or 4 quarter CM cases, respectively, diagnosed concurrently. The daily hazard or risk of diagnosis of CM in early lactation (Figure 2) was higher (P < 0.01) in heifers than in greater parity cows, and from inspection of the confidence intervals for the curves (not shown), this period of greater risk occurred from d 1 to 7 of lactation, particularly on the day following calving when the risk in heifers was approximately 3 times that of older cows. A log-rank test confirmed that survival curves to first case of CM were significantly different between groups (P < 0.01). One hundred heifers had both samples for CM and routine postcalving sampling taken at the same visit. Of the 83 heifers with 1 quarter diagnosed with CM, 41 (50%) had 1 or more additional quarters with any pathogen IMI, and of the 15 heifers diagnosed with CM in 2 quarters, 5 (38%) had 1 or more additional quarters with IMI.
|
Clustering of IMI and CM at Heifer and Herd Levels
The estimated ICC for heifer level outcomes were, in each case, greater than for those at the herd level (Table 5). The heifer-level ICC estimates were <0.20 and significantly different from 0, except for precalving infection with minor pathogens (ICC = 0.26), but significant herd-level clustering existed for major pathogen IMI postcalving and CM.
|
. Ten quarters were completely nonfunctional, and these were included with reduced function quarters for analysis because of their low prevalence. The risk of reduced function was 13% higher (P < 0.01) in quarters with, than without, major pathogens isolated postcalving, and 17% higher (P < 0.01) in quarters diagnosed with, than without, CM within 14 d following calving (IRR of 9.4 and 9.1, respectively). The risk of thelitis was significantly higher (P < 0.01) in quarters with a major pathogen isolated postcalving (incidence risk ratio = 2.3), but was similar in quarters with CM compared with normal quarters. The overall incidence of thelitis was low (<2% of teats).
|
). Other measures of milk yield and milk solids production were not significantly associated with postcalving bacteriological or CM status. Isolation of major pathogens or any pathogens postcalving in one or more quarters, or CM within 14 d of calving were significantly associated with geometric mean first herd-test SCC, equivalent to increases of 22 x 103 (P < 0.01), 17 x 103 (P < 0.01), and 12 x 103 cells/mL (P = 0.02), respectively. Isolation of a minor pathogen alone postcalving from one or more quarters was not significantly associated with ISCC. Culture of major pathogens from one or more quarters postcalving was associated with an increased risk (P < 0.05) of heifer ISCC >200 x 103 cells/mL throughout lactation (IRR = 1.95, CI = 1.65 to 2.32, incidence risk = 0.152 vs. 0.078) and at test 1 (IRR = 2.40, CI = 1.28 to 4.48), test 2 (IRR = 2.86, CI = 1.63 to 5.03), test 3 (IRR = 2.96, CI = 1.67 to 5.25), and test 4 (IRR = 1.37, CI = 1.01 to 1.87). Diagnosis of CM in one or more quarters within 14 d of calving was associated with an increased risk (P < 0.01) of heifer ISCC >200 x 103 cells/mL at test 1 only (IRR = 2.36, CI = 1.32 to 4.21). Culture of minor pathogens postcalving from one or more quarters was not associated with risk of heifer ISCC >200 x 103 cells/mL at any subsequent herd test.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
An important finding of this study is that Strep. uberis was by far the most common major pathogen causing IMI both pre- and postpartum and CM in dairy heifers in New Zealand. Streptococcus uberis was isolated from 72 and 87% of major pre- and postcalving infections, and 65% of CM cases. Precalving Strep. uberis prevalence of 2.8% of quarters was slightly lower than the 3.4% reported by Oliver and Sordillo (1988), and similar to the 2% found by Aarestrup and Jensen (1997). Still, postcalving Strep. uberis quarter level prevalence increased almost 4-fold to 10%, whereas others have reported relative increases of 2.2-fold (Oliver and Sordillo, 1988) and 1.1-fold (Aarestrup and Jensen, 1997). Reasons for the high relative importance of Strep. uberis are not known, but were suggested to be associated with pasture-grazing (Olde Riekerink et al., 2007) and may be related to the high exposure to this bacteria (Lopez-Benavides et al., 2005). The finding of a high postcalving Strep. uberis prevalence is consistent with other studies in terms of relative importance, but is much higher than the prevalence reported in a similar population by Pankey et al. (1996). Streptococcus uberis is the most common isolate from IMI and CM in all-parity cows early postpartum in New Zealand dairy systems (McDougall, 1998), suggesting a high level of exposure in all parity groups under the pasture grazing systems common in New Zealand.
Comparing quarter infection prevalence over time allowed description of the period of highest infection rate. Approximately 80% of new major IMI occurred in the last 2 to 3 wk of gestation. Major IMI prevalence did not differ significantly over the precalving study period, was maximal on the day of calving, and then declined rapidly from approximately 20 to <7% within 5 d postcalving. Therefore, the majority of new Strep. uberis infections would be occurring in the final 2 wk of gestation in pasture-grazed dairy heifers. This was not likely due to calendar date as the calving dates varied from July to September. Thus, the increase in incidence of new infection appears related to proximity to date of calving, not some climatic effect through this period.
Data from repeated samplings over time and of all quarters added important information on the patterns of IMI and CM over the peripartum period. The relatively high proportion (33% for CNS and 59% for Strep. uberis; Table 3) of precalving IMI isolated a second time in early lactation shows the importance of controlling these infections before they cause problems at calving. Over the first 5 d of lactation, only 17% of all IMI showed clinical changes, meaning that for every CM case, about 6 other IMI were undiagnosed. Half of the heifers with CM in 1 quarter had additionally 1 or more other quarters with a subclinical IMI. Failure to diagnose subclinical infections may mean that milk from these quarters will be of lower quality for processing, may increase the risk of violations of SCC standards, and may transmit infection to other quarters and animals in the herd. Only a small proportion (10%) of subclinically infected quarters within 5 d after calving was diagnosed with CM 6 to 14 d postpartum. A proportion of these subclinical IMI may have persisted into lactation, as suggested by the higher risk of ISCC >200,000 following major pathogen IMI, but most IMI did not lead to CM. Almost all the quarters cases of CM 6 to 14 d postcalving had the same isolate 0 to 5 d postcalving, reinforcing the importance of controlling new infections at or immediately before the time of calving to reduce CM and improve milk quality in later lactation.
The cumulative incidence of peripartum CM differed from those described in other countries and management systems. The 23% cumulative incidence of CM in heifers within 14 d following calving observed was higher than the 12% of heifers reported by Barnouin and Chassagne (2001), and the 8.1% of heifers reported by Pankey et al. (1996) within 5 d following calving. The finding that 7.7% of quarters with CM isolated CNS species alone is similar to that reported by Pankey et al. (1996). Although CNS are regarded as minor pathogens (Timms and Schultz, 1987) that rarely cause clinical signs or substantial increases in ISCC, they should not be disregarded as potential causes of CM in peripartum heifers. The finding of a higher hazard of early postpartum CM in heifers compared with cows supports Barkema et al. (1998). This age-related difference suggests that different risk factors for CM operate for heifers and that mastitis in heifers may have to be addressed with alternative prevention and control efforts to those against mastitis in cows.
There were differences in the distribution of IMI and CM between front and rear quarters and the pattern over time, which may have implications for their etiology. Because minor pathogens (principally CNS) were isolated in equal proportions between front and rear quarters (Table 4), these observations support the hypothesis that these are opportunistic skin pathogens (Harmon and Langlois, 1989) that are present in similar concentrations on all teats and only invade glands with open teat canals. In contrast, major environmental pathogens (principally Strep. uberis) preferentially infected rear quarters of heifers postcalving, supporting the findings of Barkema et al. (1997), and those of McDougall (1998) for CM cases in all age groups in early lactation.
Staphylococcus aureus is usually considered a contagious pathogen spread between cows in lactation during the milking process. Nevertheless, this classification does not fit our finding of Staph. aureus IMI before the first milking in heifers. Others (Roberson et al., 1994) have drawn attention to the risk of introduction of Staph. aureus to the adult milking herd by heifers with infections on multiple body sites (including teat skin and external orifices). Hence, biosecurity measures should be considered before introduction of purchased or self-sourced replacements to pasture-grazed herds in which control programs for Staph. aureus are in place.
Data from this study show that 38% of pasture-grazed heifers had one or more quarters with IMI in the pre-calving sampling period, which is much lower than the 97% found by Trinidad et al. (1990). Although caution must be taken in comparing bacteriological results from different studies due to differing methodologies, the limited available data suggest that the prevalence of IMI precalving in pasture-grazed dairy heifers in New Zealand is relatively low compared with heifers in other production systems. Nevertheless, a prevalence of 38% is high in absolute terms, suggesting significant exposure to pathogens before milking and a need for specific heifer control programs.
Herds varied numerically in their prevalence of IMI and incidence of CM, despite all using similar pasture-grazing management systems. Moreover, because an aim of this study was to describe quarter- and heifer-level, and not herd-level, patterns of IMI and CM, insufficient herds and in some cases heifers within herds were enrolled to make precise estimates of herd level measures. The estimates of ICC between quarters showed low to moderate clustering of IMI and CM measures (Table 5) at heifer level and small significant clustering at the herd level for major pathogen IMI postcalving and CM. These values were similar to those found by Barkema et al. (1997) for cows in lactation. For each outcome measure, clustering at the heifer level was several-fold greater than at the herd level. This may be interpreted that successful interventions targeting the individual animal are likely more rewarding than those aimed at herd-level management. Yet, with increasing herd sizes in New Zealand, opportunities for individual animal interventions are few, and disease control programs must usually operate at the group level. Hence, new studies are required to understand herd-level risk factors for heifer IMI and CM under pasture-based systems.
Estimation of the impact of IMI and CM on heifer productivity and longevity are important for economic analysis of proposed preventive programs. Many factors must be considered in such analyses, but important components are the effects of disease on production and longevity in the herd. Infection of a quarter with a major pathogen postcalving increased the risk 9-fold of that same quarter having reduced function subsequently (Table 6). This agrees with the finding of Waage et al. (2000), who found 25% of quarters with CM, subsequently had one or more nonfunctional quarters (19% in this study), although a feature of many cases in their study was infection with Arcanobacter spp. and Staph. aureus. Both of these are known to cause chronic and sometimes severely damaging infections, but were found to be absent or rare, respectively, in this study. Similar to their findings, thelitis occurred with higher prevalence in quarters with a major pathogen isolated postcalving. Although quarters with reduced function, asymmetry, or thelitis did not have direct economic consequences measured, these quarters are unsightly and may increase the difficulty of managing affected heifers.
Estimation of the effects of IMI and CM on milk production is not straightforward. A central problem is the confounding effect of milk production on risk of mastitis (Grohn et al., 2004). Cows with higher milk production potential may be at higher risk of mastitis than lower producing cows; hence, milk production measured in the season of CM may not be measurably different from non-CM herd mates after adjusting for other confounders such as age, breed, and days of lactation. In multiparous cows, the previous season’s milk production can be used as a covariate to control for this effect, and the estimations made within-cow and between lactations, or production from earlier in lactation can be used as a covariate. However, this is not possible in heifers in their first lactation and when mastitis occurs very early in lactation before the first herd test.
In this study, both milk yield and milk solids production recorded at the first herd test and averaged over the total lactation were used as outcome measures to estimate effect of IMI or CM on productivity (Table 7). Although the frequency of testing in New Zealand herds is low (up to 4 tests per lactation), meaning that estimates of production were imprecise, statistically significant associations may not have been detected. Data from this study found a small but significant positive association between IMI due to a minor pathogen post-calving and milk yield at the first herd test and averaged over the lactation, after adjustment for known confounders. This finding should not be interpreted as meaning that mastitis increased milk production, but more likely supports the belief that heifers with minor pathogen IMI precalving tend to have higher milk production potential. Grohn et al. (2004) found that Strep. uberis CM cases in heifers in the first week of lactation caused relatively small and short-term losses in milk production. This supports the finding of this study of no significant decrease in production in heifers with CM postcalving, when Strep. uberis was the predominant major pathogen.
Although heifers that had CM or IMI, or both, with a major pathogen or any pathogen early postpartum had significantly increased mean ISCC at first herd test or averaged over the lactation, the increase over the entire season was small. Nonetheless, significant associations existed between heifers with postcalving IMI due to major pathogens or heifers with CM, and subsequent high ISCC (>200 x 103 cells/mL). There was significant risk of high ISCC at only the first test following CM cases, whereas ISCC remained high at all tests following IMI with major pathogens postcalving. Myllys and Rautala (1995) reported higher ISCC only at the first herd test following CM cases, and not at subsequent tests in the lactation. The findings suggests that CM may have short-term effects on ISCC in heifers treated with systemic antibiotics, but that untreated subclinical IMI with major pathogens may lead to IMI that persist through the lactation and cause reduced milk quality for that period.
Replacement of cows prematurely removed from the milking herd is costly to the producer, hence, estimating the risk of removal associated with mastitis is important. The finding that heifers with major pathogens isolated 0 to 5 d after calving had a 60% increased risk of removal from the herd during lactation is important. In this study, IMI with major pathogens was diagnosed in one-third of heifers; thus, the population impact of this infection on culling is high. A possible biological reason for this finding is that CM in the postpartum period was associated with inferior subsequent reproductive performance (Chebel et al., 2004).
This study is the first to report on the epidemiology of environmental mastitis, particularly that caused by Strep. uberis, in pasture-grazed dairy heifers before and immediately following parturition. It provides information on patterns of IMI in heifers grazed under these management systems and a basis for formulation of preventive programs.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Current mastitis control programs targeting infectious pathogens are not specifically designed for heifer peripartum mastitis. They are unlikely to be successful because the environment, and not other cows, is the reservoir of the major pathogens involved, and new infections are likely occurring before the first milking when existing detection and control measures can be implemented. Novel control programs that reduce new infections due to Strep. uberis immediately before calving are required to reduce the incidence of CM in pasture-grazed dairy heifers. Effective control of environmental mastitis apart from the use of intramammary antibiotics and internal teat sealants has not been consistently reported, but because of the preponderance of one bacterial species (Strep. uberis), effective preventive measures against this organism (e.g., by vaccination) are likely to have a large population impact on CM in dairy heifers grazed on pasture.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Received for publication December 21, 2006. Accepted for publication May 16, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. W. R. Compton, C. Heuer, K. Parker, and S. McDougall Risk Factors for Peripartum Mastitis in Pasture-Grazed Dairy Heifers J Dairy Sci, September 1, 2007; 90(9): 4171 - 4180. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
—) and minor (—
—) pathogens (A) precalving and (B) postcalving, in 708 pasture-grazed dairy heifers.